OOXXXXOO / WSNet

The Computer Vision Research Toolkit

WSNets is beta version fast deeplearning training framework base on pytorch 1.3.

This project work for train, validation, test, with general dataset by different model or use the custom dataset that generate by specified data structure.

The major purpose of this project :

• Start a fast training instance on the standard datasets with default pre-trained model.
• Automatic build the dataset from original images & labels with the specified data structure rule.
• Support for some state-of-the-art models.
• Use Json config file configure the training / validation /inference instance to avoid change the project code.

---

Main Features:

1.DataSet Generator

• Automatic build the dataset from original images & labels with the specified data structure rule. In addition, it use the strict data format to help people fix the format error.

2.Fast traing from template config file

• Use Json config file configure the training / validation /inference instance to avoid change the project code.

3.Instance like training management

• Start a fast training instance on the standard or generated datasets with default pre-trained model.

4.Great automatic process & visulization

• Config decode will show up on terminal & process info will add to tensorboard where you could see the perference of training processing.

The important toolkit

Utils

• STF (Smart transform module)
• Neural Network Analysis module
• Evaluator

Neural Network Module

• PointRend
• DensePose
• CSRC { BN,DC,mask,nms,roi_align,rotated_box,wrappers }

DataSet

• temdet
• labelme2coco
• patch generate
• SatelliteData

Project Design

Description

Update schedule

Document / 文档:

EN:

---

Installation	Guide to Start	DataSet Toolkit	NetWork	DataSet	Transform	Config

---

中文:

---

安装	快速开始	数据集工具	神经网络	数据集	数据变换	配置

---

Support Mission Table:

	Segmentation	Detection	Instance Segmentation
COCO	√	√	√
NPY	√	√	√
CitysCapes	√
Pascal VOC		√
CustomDataSet	√	√	√
Open Image		√

Copyright 2020 winshare

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.